This invention relates to the field of high-energy, short-pulse generators, particularly to generators using pulsed electron beams to produce subnanosecond pulses of electromagnetic radiation.
High energy, subnanosecond electromagnetic pulses can be used for impulse, or baseband radar. They can also be used to gate electro-optic devices such as photocathodes and microchannel plates.
Such pulses are often produced directly in waveguides or on transmission lines in order to avoid lumped impedances. One such approach is to use a laser-triggered switch in a waveguide as described, for example, by G. Mourou in U.S. Pat. No. 4,329,686, issued in 1982. A large, expensive, pulsed laser is required. A second approach is to generate sharply-rising pulses using avalanche transistors or planar triodes, and then differentiate them using avalanche diodes. Pulse generators using this method are commercially-available. They have limited lifetimes, however. Moreover, the pulses are produced on coaxial transmission lines, which are unsuitable for driving loads that have non-coaxial symmetry, or require balanced signals.
An approach which has both economic and technical advantages is to make a pulse generator in which an intense pulsed electron beam is directed across a waveguiding structure comprising a waveguide or a transmission line. This approach was used in an experiment done by C. B. Norris and described by him in the Journal of Applied Physics, vol. 46, p. 1966, in 1975. In this experiment, a sheet electron beam was repeatedly swept along a stripline containing a semiconductor dielectric. Electron amplification was provided by the semiconductor. The beam produced electromagnetic pulses in the stripline. The pulses were made short and intense by synchronizing the velocity of the beam along the stripline with that of each pulse. The synchronization was similar to that which occurs when an electromagnetic pulse is produced by a nuclear weapon exploded just outside the earth's atmosphere. This approach was limited in practicality by the low charge per pulse available from the beam, by the low amplification of the semiconductor, and by the large size of the apparatus.
The charge can be increased by using a larger cathode. Instead of sweeping the beam, the cathode can be pulsed and a constant high voltage can be used to accelerate the electrons. A large, pulsed cathode can be made by using a pulsed laser to illuminate a photocathode in vacuum. This method was described by M. T Wilson and P. J. Tallerico in U.S. Pat. No. 4,313,072, issued in 1982. The pulsed laser has the above-mentioned disadvantages, however. Moreover, a photocathode requires an ultra-high vacuum in which to operate, and even then degrades rapidly under high-current conditions.
A pulsed cathode can also be mad by using a triode which contains a thermionic or a field-em cathode. The triode requires high-power trigger pulses, however. The trigger pulses must have durations less than or equal to those of the electron pulses that are to be generated. They are usually produced with considerable jitter.
To obtain a pulsed cathode without the above drawbacks, it can be constructed in two stages. In the first stage, low-power, low-jitter trigger pulses are used to produce a reliable, low-intensity pulsed electron beam that can cover a substantial area. In the second stage, an electron multiplier (EM) with a large area is used to amplify the beam. The amplification possible depends upon the duty cycle and area of the EM. When used intermittently, EMs can produce pulses with large electron charge densities, on the order of one nanocoulomb per cm.sup.2. The total charge increases with the area of the EM. When accelerated by voltages in ,the range of tens of kilovolts, such electron pulses can attain high kinetic energies, on the order of tens of microjoules. Large electromagnetic pulses can be produced from these electron pulses, even if the energy-conversion efficiency is small.
One EM which is occasionally used indirectly to amplify broad-area electron beams is the microchannel plate (MCP). Amplification of such beams has been seen to occur in the course of measurements of the intensities of brief flashes of light. MCP photomultiplier tubes (PMTs) have been flooded with flashes, and electromagnetic pulses have been produced as a result. Examples of such measurements are provided in an article by L. P. Hocker et al. in the IEEE Transactions on Nuclear Science, vol. 26, p. 356, published in 1979. The MCP PMTs used for these measurements were not optimally-suited for producing electromagnetic pulses, however. Due to the coaxial symmetry of these devices, there was an inherent conflict between providing enough accelerating voltage, and providing impedance matching within the PMTs. Even if these problems could be solved, the resulting pulses produced would still be restricted to being negative and unbalanced, and the pulse durations would be limited by the widths of the anodes in the PMTs.